Quantitative legal prediction as a dispute resolution mechanism

Abstract

AI developments and their potential impact on society have dominated the public debate in recent years. This has not been different in the legal field, where many hope that AI could bring about disruptive change in an area of society that has until today proven to be largely insulated against technological developments. The legal system produces a massive amount of data (legal decisions, lawyers’ briefs, etc.), which, through the use of AI techniques such as natural language processing and machine learning, could be harnessed and instrumentalized for several different purposes. The image that has captured the imagination is that of the robot judge, dispending swift and impartial justice without human intervention. Legal conflicts, however, decide the lives and livelihoods of people, and these high stakes explain why the administration of justice is surrounded by fundamental procedural rights. The robot judge is today far from reality, but AI is increasingly finding its way into the legal system and its use as a (crude) dispute resolution mechanism is no longer unimaginable. This contribution discusses the use of AI systems within the legal field as a means for dispute resolution. It briefly touches upon existing applications and the possible advantages, but focuses mainly on the many challenges that the interplay between AI and dispute resolution poses in light of fundamental procedural rights. It concludes with a discussion of the European regulatory response to this new phenomenon.